简介：Thispaperproposesanovelexemplar-basedmethodforreducingnoiseincomputedtomography(CT)images.Intheproposedmethod,denoisingisperformedoneachblockwiththehelpofagivendatabaseofstandardimageblocks.Foreachnoisyblock,itsdenoisedversionisthebestsparsepositivelinearcombinationoftheblocksinthedatabase.Weformulatetheproblemasaconstrainedoptimizationproblemsuchthatthesolutionisthedenoisedblock.Experimentalresultsdemonstratethegoodperformanceoftheproposedmethodovercurrentstate-of-the-artdenoisingmethods,intermsofbothobjectiveandsubjectiveevaluations.

简介：Inordertoovercomethedisadvantagesoflowaccuracyrate,highcomplexityandpoorrobustnesstoimagenoiseinmanytraditionalalgorithmsofcloudimagedetection,thispaperproposedanovelalgorithmonthebasisofMarkovRandomField(MRF)modeling.Thispaperfirstdefinedalgorithmmodelandderivedthecorefactorsaffectingtheperformanceofthealgorithm,andthen,thesolvingofthisalgorithmwasobtainedbytheuseofBeliefPropagation(BP)algorithmandIteratedConditionalModes(ICM)algorithm.Finally,experimentsindicatethatthisalgorithmforthecloudimagedetectionhashigheraverageaccuracyratewhichisabout98.76%andtheaverageresultcanalsoreach96.92%fordifferenttypeofimagenoise.

简介：全部的变化(电视)最小化问题广泛地在图象恢复被学习。尽管许多其他的方法为它的答案被建议了，牛顿方法由于没有集中为最初的明确的表达仍然保持不可用。由禅宗，周和禅宗的以前的研究[15]认为一个规则化参数继续想法与一些成功而是没有柔韧的参数选择增加牛顿方法的集中的领域计划。在这份报纸，我们为一样的最初的电视明确的表达考虑一个homotopy方法并且建议使用曲线追踪适应地选择规则化参数。结果，这个想法帮助实质地在高效地解决TVEuler-Lagrange方程改进以前的工作。一样的想法也象deblurring问题一样为二个另外的方法被考虑，再与获得的改进。数字实验证明我们的新方法为图象恢复柔韧、快，甚至为有大noisy-to-signal比率的图象。

简介：纸为同时的图象分类和注解建议一个新奇概率的生产模型。模型考虑范畴信息能为图象注解提供珍贵信息的事实。一旦一幅图象的范畴被查明，注解词的范围能被缩小，并且产生无关的注解词的概率能被减少。到这个目的，根据班注解图象的想法在模型被介绍。用变化方法，模型的近似推理和参数评价算法被导出，并且为分类并且注解新图象的有效近似也被给。我们的模型的力量在二真实世界数据集上被表明：1600图象LabelMe数据集和1791图象UIUC运动数据集。实验结果证明分类表演与几个最先进的分类模型同样，当注解表演比几个最先进的注解模型的好时。

简介：Weproposeanewalgorithmforthetotalvariationbasedonimagedenoisingproblem.ThesplitBregmanmethodisusedtoconvertanunconstrainedminimizationdenoisingproblemtoalinearsystemintheouteriteration.Analgebraicmulti-gridmethodisappliedtosolvethelinearsystemintheinneriteration.Furthermore,Krylovsubspaceaccelerationisadoptedtoimproveconvergenceintheouteriteration.Numericalexperimentsdemonstratethatthisalgorithmisefficientevenforimageswithlargesignal-to-noiseratio.

简介：MostofImageQualityAssessment(IQA)metricsconsistoftwoprocesses.Inthefirstprocess,qualitymapofimageismeasuredlocally.Inthesecondprocess,thelastqualityscoreisconvertedfromthequalitymapbyusingthepoolingstrategy.Thefirstprocesshadbeenmadeeffectiveandsignificantprogresses,whilethesecondprocesswasalwaysdoneinsimpleways.Inthesecondprocessofthepoolingstrategy,theoptimalperceptualpoolingweightsshouldbedeterminedandcomputedaccordingtoHumanVisualSystem(HVS).Thus,areliablespatialpoolingmathematicalmodelbasedonHVSisanimportantissueworthyofstudy.Inthispaper,anewVisualPerceptualPoolingStrategy(VPPS)forIQAispresentedbasedoncontrastsensitivityandluminancesensitivityofHVS.ExperimentalresultswiththeLIVEdatabaseshowthatthevisualperceptualweights,obtainedbytheproposedpoolingstrategy,caneffectivelyandsignificantlyimprovetheperformancesoftheIQAmetricswithMeanStructuralSIMilarity(MSSIM)orPhaseQuantizationCode(PQC).ItisconfirmedthattheproposedVPPSdemonstratespromisingresultsforimprovingtheperformancesofexistingIQAmetrics.

简介：Thispaperfocusesontheproblemofautomaticimageclassification(AIC)byproposingaframeworkbasedonlatentsemanticanalysis(LSA)andimageregionpairs.Thenovelframeworkemploysrelativespatialarrangementsforregionpairsastheprimaryfeaturetocapturesemantics.Thesignificanceofthispaperistwofold.Firstly,tothebestourknowledge,thisisthefirststudyoftheinfluenceofregionpairsaswellastheirrelativespatialinformationinlatentsemanticanalysisasappliedtoautomaticimageclassification.Secondly,ourproposedmethodforusingtherelativespatialinformationofregionpairsshowgreatpromiseinimprovingimagesemanticclassificationcomparedwiththeclassicallatentsemanticanalysismethodand2Dstringrepresentationalgorithm.

简介：摘要计数阳性细胞是Image-ProlusP的常见应用。然而由于原始照片质量不佳或拍摄视野不能覆盖观察区域等问题，用Image-ProPlus计数阳性细胞可能得不到理想的结果甚至失败。通过提高原始照片质量、拼图及不同计数方法的应用能有效应对这些问题。

简介：Imagedenoisingisthebasicproblemofimageprocessing.Quaternionwavelettransformisanewkindofmultiresolutionanalysistools.Imageviaquaternionwavelettransform,waveletcoefficientsbothinintrascaleandininterscalehavecertaincorrelations.First,accordingtothecorrelationofquaternionwaveletcoefficientsininterscale,non-Gaussiandistributionmodelisusedtomodelitscorrelations,andthecoefficientsaredividedintoimportantandunimportancecoefficients.Thenweusethenon-Gaussiandistributionmodeltomodeltheimportantcoefficientsanditsadjacentcoefficients,andutilizetheMAPmethodestimateoriginalimagewaveletcoefficientsfromnoisycoefficients,soastoachievethepurposeofdenoising.Experimentalresultsshowthatouralgorithmoutperformstheotherclassicalalgorithmsinpeaksignal-to-noiseratioandvisualquality.

简介：为打印应用程序的工业大3D的一个新奇7轴机器人在这份报纸被介绍，它被机器人学和自动信息系统的研究所开发，Nankai大学。原型上的底，三维的打印上的机制深深地被把。由在打印的3D调查颜色失真的原因，一个新奇multistep赔偿算法基于模型优化和图象赔偿被开发。最后，试验性的结果显示出打印机器人站台和建议算法的3D的表演。

简介：Aninfraredimagedetailenhancementmethodbasedonlocaladaptivegammacorrection(LAGC)isproposed.ThelocaladaptivegammavaluesaredesignedbasedontheWebercurvetoenhanceeffectivelytheimagedetails.Subsequently,theactivegrayscalerangeoftheimageprocessedbyLAGCisfurtherextendedbyusingourproposedhistogramstatisticalstretching.Theexperimentalresultsshowthattheproposedalgorithmcouldconsiderablyincreasetheimagedetailsandimprovethecontrastoftheentireimage.Thus,ithassignificantpotentialforpracticalapplications.

简介：把复杂contourlet变换(CCT)与nonnegative矩阵因式分解(NMF)相结合的一个图象熔化方法在这份报纸被建议。在二幅图象被CCT分解以后，NMF被用于他们的highand低频率的部件分别地，一幅图象被综合。图象熔化结果的Subjective-visual-quality基于NMF和有NMF的小浪/contourlet/nonsubsampledcontourlet的联合与图象熔化方法的那些相比。试验性的结果是评估份量上，并且跑的时间也被对比。建议图象熔化方法能获得更大的信息熵，标准差和吝啬的坡度，这被显示出，它意味着它能更好从所有来源图象集成展示信息，避免背景噪音并且有效地支持在熔化图象的空间明白。

简介：Aimingatfacilitatingtheresearchofurbantourismimagepositioning,thispapersummarizesthedomesticandabroadtheoriesonurbantourismimageandanalyzesitssignificanceforcities.WithDengfengasanexample,itsgoalistoboostdevelopmentofDengfengtourism,whichisanexcellenttourismcityinChina.Thispaperpresentsitscurrenturbantourismdevelopingsituation,positionsitstourismimage,concludestheplanofimagebrandandproposespromotionalslogansbasedontheanalysisofDengfengtourismlocalfeatures,itscoreelementsandperceptionsofvisitors.

简介：Thisstudywasonsuperiorityofthenon-negativematrixfactorization（NMF）algorithmforapplicationofinformationextractedwithaerialimages.First,NMFwasusedforaerialimageinformationextraction,andthenthisdatawascomparedwithaprincipalcomponentanalysis（PCA）inwhichr（thenumberofrowsorcolumnsofbasicmatrix）andEignum（thenumberofeigenvalues）weregivendifferentvalues.ExperimentalresultsshowedthattheruntimeofNMFwithr=20or50waslessthanthatofPCAwithanEignum=20or50.Also,therecognitionrateofNMFwithr=50washigherthanthatofanEignum=50.Theexperimentshowedthatnonnegativematrixfactorizationhadadvantagesofashorttimeperiodwithahighrecognitionrate.

简介：Anovelactivecontourmodelisproposed,whichincorporateslocalinformationdistributionsinafuzzyenergyfunctiontoeffectivelydealwiththeintensityinhomogeneity.Moreover,theproposedmodelisconvexwithrespecttothevariablewhichisusedforextractingthecontour.Thismakesthemodelindependentontheinitialconditionandsuitableforanautomaticsegmentation.Furthermore,theenergyfunctionisminimizedinacomputationallyefficientwaybycalculatingthefuzzyenergyalterationsdirectly.Experimentsarecarriedouttoprovetheperformanceoftheproposedmodeloversomeexistingmethods.Theobtainedresultsconfirmtheefficiencyofthemethod.

简介：Inthispaper,elitistreconstructiongeneticalgorithm(ERGA)basedonMarkovrandomfield(MRF)isintroducedforimagesegmentation.Inthisalgorithm,apopulationofpossiblesolutionsismaintainedateverygeneration,andforeachsolutionafitnessvalueiscalculatedaccordingtoafitnessfunction,whichisconstructedbasedontheMRFpotentialfunctionaccordingtoMetropolisfunctionandBayesianframework.Aftertheimprovedselection,crossoverandmutation,anelitistindividualisrestructuredbasedonthestrategyofrestructuringelitist.ThisprocedureisprocessedtoselectthelocationthatdenotesthelargestMRFpotentialfunctionvalueinthesamelocationofallindividuals.Thealgorithmisstoppedwhenthechangeoffitnessfunctionsbetweentwosequentgenerationsislessthanaspecifiedvalue.Experimentsshowthattheperformanceofthehybridalgorithmisbetterthanthatofsometraditionalalgorithms.

简介：异构或合成的材料的有效热传导性为在科学和工程的特定的功能是材料选择和设计的一个必要物理参数。有效热传导性重重地在部分和每个阶段的空间分发上被依靠。在这个工作，基于图象的有限元素方法(女性)被用来与不同毛孔结构计算多孔的陶艺的有效热传导性。与象理论(EMT)方程和平行建模的有效媒介那样的以前的理论模型相比，基于图象的女性的罐头与相对稳定的偏差被用于材料系统的一个大变化。基于图象的女性的计算的偏差主要来自之间的差别二维(2D)想象并且三维(3D)真实系统的结构，和一个实验被执行证实这个假设。因素影响2D和3D有效热传导性被学习由对女性说明精确性和申请条件基于图象女性。